I've been looking back to try and pick up the context, but are we talking about the LED/zener reference used to bias the upper mosfet in the shunt regulator? If so, the R's and C's can be shifted around somewhat for more effective filtering of of whatever voltage reference is used, and may make it possible to use the transistor junction zener without sonic penalty. Just a thought.
No, not in the reg. There is a local basic shunt (a CCSed and cap bypassed Zener) creating base bias for the cascode bjt/jfet first stage in the RIAA.
Since the Zener was one 8V2 part there, we like to know if we will listen better either with a LEDs string or 2 low voltage Zeners. Its unfortunate that our 7.5-8.5V ref need, just coincides with the worst single Zener values for noise. Under 4V7 or 12V and over, single Zeners are great for self noise. Especially if someone has access to temp compensated parts reserved for instrumentation and aerospace.
Since the Zener was one 8V2 part there, we like to know if we will listen better either with a LEDs string or 2 low voltage Zeners. Its unfortunate that our 7.5-8.5V ref need, just coincides with the worst single Zener values for noise. Under 4V7 or 12V and over, single Zeners are great for self noise. Especially if someone has access to temp compensated parts reserved for instrumentation and aerospace.
It might also be a good idea to check several types of transistors for noise, as I would bet that it varies with device geometry. A string of yellow-green LEDs is a relatively no-fuss way to go, though. I hate low voltage zeners on principle, as it takes a lot of current to bias them into their sweet spot.
Post #1797 shows 5mA good, 20mA best for noise. That was for 2 samples of a 3V9Z. 10mA IDSS CCS can do it decently I guess, based on that data.
On the other hand 4X yellow ones (red come 2nd, green 3rd for noise, slight differences, but green stay better grouped for individual Vdrop), can be used, especially if miniature or surface mount ones, to help with Ricardo's tight space thing.
On the other hand 4X yellow ones (red come 2nd, green 3rd for noise, slight differences, but green stay better grouped for individual Vdrop), can be used, especially if miniature or surface mount ones, to help with Ricardo's tight space thing.
I am starting to lean towards leds but meanwhyle I just found some zeners BZX79C3V6 that fix on 3.8v so I can get 7.6v with two of those. (Can I use 0.5w zeners ?)
Can I expect some improvements by using zeners in paralel ? (I could get 1w that way)
Can I expect some improvements by using zeners in paralel ? (I could get 1w that way)
Yes you can use 0.5W. Using Zeners in parallel will lower their intrinsic noise (which is good enough at those voltages anyway), but they will share the available current so you are back to square 1. If you had double the current it would be kind of meaningful.
a 500mW 3v6 Zener has a maximum current of 139mA.
Use >=13.9mA when trying to get a voltage reference from it.
Do not use Zeners in parallel.
Adding two 3v6 Zeners in series to get 7.2V still requires >=13.9mA.
4 red LEDs in series will do much better with just 2 to 3mA.
Use >=13.9mA when trying to get a voltage reference from it.
Do not use Zeners in parallel.
Adding two 3v6 Zeners in series to get 7.2V still requires >=13.9mA.
4 red LEDs in series will do much better with just 2 to 3mA.
D1 modification
Gentlemen....after reading your comments, I decided to use four miniature yellow leds.
With 10mA (From jfet setup), I got 7.65v.
After some dreadfull desoldering, resoldering, the riaa finally "lit up".
Imediate gains are noticeable in the noise floor....
Bass is much more inteligible with greater detail and speed.
Trebble is sweet and perfectly integrated.
I noticed no change in the instruments "tone".
Must listen for a while to be able to report correctly, but I believe this is a "must" do mod.
Ricardo
Gentlemen....after reading your comments, I decided to use four miniature yellow leds.
With 10mA (From jfet setup), I got 7.65v.
After some dreadfull desoldering, resoldering, the riaa finally "lit up".
Imediate gains are noticeable in the noise floor....
Bass is much more inteligible with greater detail and speed.
Trebble is sweet and perfectly integrated.
I noticed no change in the instruments "tone".
Must listen for a while to be able to report correctly, but I believe this is a "must" do mod.
Ricardo
OK! That's encouraging. I was smelling a rat with the cascode version's local bias... Thought it could be done better. Listen more and if everything is positive, I will post 3 final schematics. MM/highMC, Medium MC, MC, with LEDs.
P.S. Do you have a ''yellow submarine'' RIAA photo?
P.S. Do you have a ''yellow submarine'' RIAA photo?
Here is the final circuit in 3 sensitivities, best we know of. CloadX is additional to cable capacitance if needed for MM. See manufacturer specs. A TT normally presents 150-200pF from cart to male RCA.
For MC it is normally omitted, unless there is some recommendation from a cartridge manufacturer. Rload is the load resistor. You can start with the ballpark example values, but better fine tune. See MC cart spec again.
1. MM or High MC. 2-5mV.
For MC it is normally omitted, unless there is some recommendation from a cartridge manufacturer. Rload is the load resistor. You can start with the ballpark example values, but better fine tune. See MC cart spec again.
1. MM or High MC. 2-5mV.
Attachments
HI Ricardo , mod_man 🙂
I read that put the Rload back on head shell (non fixed) sound better ,we can put on arm pod for fixed headshell....
I read that put the Rload back on head shell (non fixed) sound better ,we can put on arm pod for fixed headshell....
What is the type and sensitivity of your speakers? What amp and preamp you have? Do you know their gain in dB? How can you characterize the sound of your system right now? Heavy, thin, or neutral?
Because I know the DL103 and the cascode or not I am trying to give you the best advice. Let me know about the questions.
Because I know the DL103 and the cascode or not I am trying to give you the best advice. Let me know about the questions.
Very good, Salas, well done.Salas said:
Now, next step, a small BOM with suggested caps (model/brand) and final shunt schematics for the different ps voltages
Am I asking too much? 😀
P.S. do you need something from Italy? I'm flying to Greece next weekend 😉
Like you knew it, I am in contact with USA member psgr and we create a resume thread pdf, with BOM's, parts suggestions etc.
Can I have a desmo bike from Bologna please?😀 Nice for summer.😎
Can I have a desmo bike from Bologna please?😀 Nice for summer.😎
